Recreation Hot-Spots - Oman Case¶

The tourism sector is a major contributor to the economies of many countries, and its success depends on a variety of factors, including the quality of ecosystems, cultural environment, and safety. Tourists often choose their destinations based on the quality and availability of services in a given area.

One way to evaluate the economic value of recreational activities and the trade-offs associated with different land use and management decisions is to use the InVEST recreation model. The model produces output datasets that can be used to identify hot-spots and areas suitable for further development on a national scale. By analyzing the spatial distribution of recreational activities and the factors that influence people's decisions to visit different areas, the model can help planners and decision-makers understand the needs and preferences of tourists and prioritize investments in recreation infrastructure and resources.

Tourism is an important contributor to the economy of Oman, and the country has a rich culture and diverse natural beauty that attract visitors from around the world.

The InVEST Recreation model¶

The InVEST recreation model is a tool that helps planners determine where visitors go and what influences their decisions to visit different sites. The decision to visit a particular site is based on many different factors, such as accessibility, cultural heritage and history, safety, and environmental factors.

The use of the InVEST model can help planners and decision-makers understand the hot recreation spots and where more visitors tend to go, and can help identify areas for further development. It can also be used to estimate the infrastructure needed to extend or develop hot spots within an area. The first step in the analysis is to investigate the spatial distribution of recreational-attraction areas and to identify touristic and recreational hot spots.

Model parameters¶

In order to identify hot-spots, the InVEST model is used to obtain critical information on visitation patterns in the area of interest. The model uses data scraped from social media (such as Flicker) as an indication of visitation rates. The model requires that the area of interest be divided into a grid of cells, and all output data are linked to this vector grid file. In the case of Oman, a 5x5 km grid cell size was used. In each grid cell, the number of person-days are recorded from 2005 to 2017:

By year average,
By moth average,
Monthly visits for the selected period.

For more information on the InVEST recreation model, you can visit the following link: https://naturalcapitalproject.stanford.edu/software/invest-models/recreation

Recreation Distribution Data¶

The InVEST recreation model produces two datasets that are used in the hot-spot delineation method:

The visitation distribution layer: This is a shapefile that divides the area of interest (in this case, the whole country of Oman) into grid cells. The cell size is defined as a parameter in the InVEST model and is set at 5 km for this study.
The monthly visitation distribution data table: This is a CSV file that contains the recorded number of person-days (visits) for all available years (2005-2017 as defined in the InVEST model). The table includes the average number of visits per year and per month, as well as the total number of monthly visits for the selected period.

MapView(layout=Layout(height='400px', width='100%'))

Cleaning the data sets¶

The InVEST recreation model divides the area of interest into grid cells and associates each cell with information about the number of visits to that area over a specified period of time. However, in many cases, especially in areas with large deserts like Oman, there may be many cells that have no recorded visits. These cells in the visitation distribution layer and the monthly distribution data table should be deleted as they provide no useful information and could potentially skew the analysis. It is important to clean the data sets by removing these cells in order to ensure that the results of the analysis are accurate and reliable.

From a total of 12665 cells only 2355 cells have recorded at least one visit between 2005 - 2017

The Recreation Distribution Layer¶

Now that the Recreation Distribution Layer is clean it can be displayed for the whole country of Oman.

MapView(layout=Layout(height='400px', width='100%'))

The map above provides a clear understanding of the areas where more people tend to visit and how hot-spot areas tend to form. These hot-spot areas will be defined in the following section. Below, the attributes of the recreation distribution are displayed. For each cell, the following information is provided:

PUD_YR_AVG: The total year visitation average. For this example, the selected time frame is from 2005-2017 as provided by the InVEST database.
PUD_mmm (where mmm is each calendar month): The monthly average for each month. Again, the time frame is from 2005-2017. This information can be used to understand the seasonal patterns of visitation and identify the months when hot-spot areas are most popular.

	FID	PUD_YR_AVG	PUD_JAN	PUD_FEB	PUD_MAR	PUD_APR	PUD_JUL	PUD_AUG	PUD_SEP	PUD_OCT	PUD_NOV	geometry
0	0	0.38	0.08	0.15	0.00	0.00	0.00	0.00	0.15	0.00	0.00	POLYGON ((83220.313 1844894.490, 83220.313 184...
3	3	0.38	0.08	0.08	0.08	0.00	0.00	0.00	0.08	0.08	0.00	POLYGON ((83220.313 1849894.490, 83220.313 185...
4	4	0.77	0.08	0.23	0.00	0.00	0.08	0.08	0.23	0.08	0.00	POLYGON ((88220.313 1849894.490, 88220.313 185...
5	5	0.62	0.00	0.00	0.00	0.00	0.08	0.15	0.23	0.15	0.00	POLYGON ((93220.313 1849894.490, 93220.313 185...
6	6	1.08	0.00	0.15	0.00	0.08	0.15	0.08	0.38	0.15	0.08	POLYGON ((98220.313 1849894.490, 98220.313 185...

The monthly distribution table:¶

The monthly distribution dataset provides information on visits to each cell and specifies the exact year and month when the visit took place. The table contains the cell ID (poly_id column) and the monthly visits in the following format: YYYY-M, where:

YYYY is the calendar year.
M is the month's sequence number (e.g., January = 1, December = 12).

The table also includes a field called "Total," which records the total number of visits to the cell from 2005-2017. This information can be used to understand the overall trends in visitation over time and identify which cells are the most popular.

	poly_id	...	Total
0	0	...	5
2	3	...	5
4	4	...	10
5	5	...	8
6	6	...	14

5 rows × 158 columns

Define Recreation Hot-Spot areas Methodology¶

Having clean data sets, it is now possible to proceed with the hot-spot area delineation. In the sections below, the method for defining the recreation hot-spot areas is described in detail. This method is based on the use of Geographic Information System (GIS) spatial statistics techniques to analyze spatial patterns, identify hot-spots, and visualize their spatial dimensions. This method is valuable for policymakers as it provides insights into the spatial distribution of tourist flows, enabling better resource allocation, management, and protection of hot-spot areas. It can also be useful for the private sector and tour operators, as it helps them to optimize the economic utilization of tourist clusters and hot-spots.

Calculate the total year average visitation rate¶

For the whole country, the total year average is calculated by adding the monthly averages.

The total visitation year average for the country of Oman is 1738.

The total visitation year average rate does not provide a quantitative measure of the number of tourists, but rather a measure of the quality of the recreational experience. This means that areas with higher visitation rates tend to receive more visitors. If the visitation rate is 50% higher in an area, it means that there are 50% more visitors at this location compared to areas with lower rates. With this in mind, it is possible to calculate the importance of each hot-spot area and the share of visitations that each site attracts. This information can be useful for policymakers and businesses as it helps to understand the demand for different recreational sites and how to allocate resources to meet this demand.

Define Hot-Spot areas¶

To define the hot-spot areas, it is necessary to define a visitation threshold value. This value is used to classify the grid cells in the following two categories:

Grid cells with a value greater than the threshold value: These cells are considered as candidate cells for hot-spot area cores because they have attracted visitors in the past and are likely to attract visitors in the future. These cells are expected to have high recreation potential and may be part of a hot-spot area, particularly the core areas.
Grid cells with a value less than the threshold value: These cells do not have the potential to attract visitors and have low recreation potential. It is also possible that the recorded number of person-days in these cells is random or even a mistake. While these cells will not automatically be included in the hot-spot areas, some of these cells may end up being part of these areas.

Method Steps¶

Spatial processing
1. Select core area cells.
2. Merge core area cells to form candidate hot-spot areas.
3. Select hot-spot area cores.
4. Extend core areas.
5. Add names to areas (optional).
Attribute processing
1. Classify cells into hot-spots and the remaining country category.
2. Aggregate visitation rates.

This method provides a systematic approach for identifying and defining hot-spot areas based on the recorded number of visitors. It allows planners to understand the spatial distribution of recreational demand and make informed decisions about how to allocate resources and develop recreational sites.

Spatial Processing¶

Initially, hot-spot areas are identified through a GIS method. While high-value recreation areas may often be identified as hot-spots based on past experiences of planners or the characteristics of those areas, GIS allows planners and government agencies to more accurately pinpoint hot-spots, confirm the high-value of these areas, identify the specific activities occurring within the hot-spot, and develop strategies for further development and management.

Selecting Core Area Cells¶

The potential hot-spot areas are the cells that have high visitation rates. All cells that have a value greater than the threshold are potential hot-spot area cells. To select these cells, it is necessary to identify all cells with a visitation value greater than the threshold. This can be done using GIS software or other spatial analysis tools. Once these cells are identified, they can be used as the core areas for defining the hot-spot areas.

Selected cells with visitation rate greater than the threashold

	FID	PUD_YR_AVG	January	February	March	April	July	August	...	Total
0	0	0.38	0.08	0.15	0.00	0.00	0.00	0.00	...	5
1	3	0.38	0.08	0.08	0.08	0.00	0.00	0.00	...	5
2	4	0.77	0.08	0.23	0.00	0.00	0.08	0.08	...	10
3	5	0.62	0.00	0.00	0.00	0.00	0.08	0.15	...	8
4	6	1.08	0.00	0.15	0.00	0.08	0.15	0.08	...	14

5 rows × 174 columns

Total number of cells with value > of threashold is 861
From the total 2355 cells with at least one recorded visit, only 861 cells are potentially core cells.

Merging Core Area Cells:¶

The selected cells are still individual units, even if they are located in the same area and are neighboring each other. By removing the common boundaries and merging the selected cells, new polygons are formed. These polygons are the candidate polygons for hot-spot areas. However, not all of these new polygons will be considered as hot-spots, as the size of the hot-spot areas is also a criteria that is introduced as a parameter in the selection process. To merge the cells, GIS software or other spatial analysis tools can be used to combine the cells into larger polygons based on the specified criteria. This process helps to identify larger areas that may have a higher recreation potential and can be considered as hot-spots.

Selecting Hot-Spot Area Cores¶

The recreation hot-spot areas should have a minimum size, as the study is on a national level and the cell size is 5km. Therefore, in the case of Oman, the minimum size for a hot-spot area is set to 200km2. Using this criterion, the cores of the hot-spot areas can be selected and displayed. This process helps to identify the areas that have the highest recreation potential and can be considered as the cores of the hot-spot areas. GIS software or other spatial analysis tools can be used to select the cells that meet the specified criteria and create the hot-spot area cores.

Display Core Hot-Spot Areas.

Make this Notebook Trusted to load map: File -> Trust Notebook

Extend core areas¶

Until now two criteria are used to define the core areas. The value of the individual cell (to be greater than the threshold) and the size of the core (to be over 200km2). But the output is fragmented and many polygons are too close and should be somehow merged. In this step, the core areas are extended by 5km with a buffer. These new polygons are the recreation final hot-spot areas.

Extending Core Areas¶

So far, two criteria have been used to define the core areas: the value of the individual cell (which must be greater than the threshold) and the size of the core (which must be over 200km2). However, the output may still be fragmented, with many polygons being too close and in need of merging. In this step, the core areas are extended by 5km with a buffer. These new polygons are the final recreation hot-spot areas. GIS software or other spatial analysis tools can be used to create the buffer around the core areas and create the final hot-spot areas. This process helps to identify larger, contiguous areas that have the highest recreation potential and can be considered as the final hot-spot areas.

Display Extended Core Hot-Spot Areas.

Make this Notebook Trusted to load map: File -> Trust Notebook

Adding Names to Hot-Spot Areas¶

This is an optional but very helpful step. By adding names to the hot-spot areas, locals can easily identify and verify the results, as they can link the hot-spot characteristics and visualize the attributes of each area. This process helps to make the hot-spot areas more understandable and accessible to the public, and can also help with communication and engagement with stakeholders and decision-makers.

Add names to Hot-spot areas.

	level_1	ID	geometry	area	Name
0	0	100	POLYGON ((83220.313 1839894.490, 82730.227 183...	5743.368709	Salalah
1	1	101	POLYGON ((663896.427 2482394.490, 663810.707 2...	2322.237082	Al Wasil
2	2	102	POLYGON ((738220.313 2464894.490, 737730.227 2...	1146.034281	Sur
3	3	103	POLYGON ((378220.313 2669894.490, 377730.227 2...	2555.169962	Al Buraimi
4	4	104	POLYGON ((448896.427 2707394.490, 448810.707 2...	1014.532253	Sohar
5	5	105	POLYGON ((408220.313 2874894.490, 407730.227 2...	2981.047366	Musandam
6	6	106	POLYGON ((520720.313 2584218.376, 520863.329 2...	5516.060452	Nizwa
7	7	107	POLYGON ((713220.313 2514894.490, 708220.313 2...	2639.857650	Dibah
8	8	108	POLYGON ((603220.313 2599894.490, 602730.227 2...	4238.752195	Muscat

Display Final Core Hot-Spot Areas.

Make this Notebook Trusted to load map: File -> Trust Notebook

Attribute Processing¶

Until now, the hot-spot areas have been spatially defined. In this section, the attributes or visitation characteristics of the hot-spot areas will be calculated. Before the delineation of the hot-spot areas, the grid cell was the only spatial unit, and all information was linked to cells. With the introduction of hot-spot areas, the attributes of cells should be aggregated and linked to the new polygons.

Classifying Cells to Hot-Spot and Remaining Country¶

Cells that fall within the hot-spot areas will pass their attributes (or aggregate information) to the respective polygon. The rest of the cells are classified as "remaining country" or "country" to estimate the importance of each hot-spot area and the significance of the remaining areas. This process helps to understand the distribution of visitor activity across the region and identify the most important hot-spot areas.

Aggregating Visitation Rates¶

The visitation rates of each cell can be aggregated and linked to the hot-spot areas to understand the overall importance of each area in terms of visitor activity. This process helps to identify the hot-spot areas that receive the highest number of visitors and the areas that may be less popular.

Cells that do not fall within the hot-spot areas will be characterized as part of the "remaining country". This is a class that contains the remaining cells that are scattered around the hot-spot areas. This class is without a defined spatial representation. The visitation values for this class are aggregated and displayed below to understand the overall importance of the remaining areas in terms of visitor activity. This process helps to identify areas that may not be considered hot-spots but still receive a significant number of visitors.

Aggregate visitation rates¶

In this step, the visitation rates for each cell within the hot-spot areas are aggregated and displayed at the polygon level for each year and month. This process allows for a better understanding of the overall visitor activity within each hot-spot area, rather than just looking at individual cells. The aggregated values provide a more comprehensive view of the visitor patterns within the hot-spot areas and can be used to inform decision making and resource allocation.

Aggregated visitation values for all hot-spot areas

		PUD_YR_AVG	January	February	March	April	May	June	July	August	September	October	November	December
ID	Name
100	Salalah	201.54	15.15	11.85	14.46	13.77	9.85	9.69	20.00	27.69	28.31	17.46	19.54	13.77
101	Al Wasil	51.54	6.77	6.85	4.08	4.08	1.92	0.38	0.62	1.00	1.08	7.23	8.77	8.77
102	Sur	41.69	5.38	4.77	5.69	3.15	1.92	1.77	1.15	2.08	1.85	4.31	4.69	4.92
103	Al Buraimi	54.46	6.85	5.62	5.92	3.62	4.08	3.15	4.23	3.31	2.92	3.85	5.54	5.38
104	Sohar	23.77	1.31	2.54	2.15	2.62	2.85	2.08	1.69	1.85	1.08	1.38	2.46	1.77
105	Musandam	107.92	9.77	13.23	13.38	8.54	7.08	4.85	4.00	5.62	5.77	9.38	12.38	13.92
106	Nizwa	209.85	24.92	31.15	19.46	19.23	9.46	6.92	8.38	4.15	9.23	17.92	26.85	32.15
107	Dibah	77.15	10.77	9.92	6.15	6.92	4.85	3.92	2.31	2.62	2.92	8.62	7.38	10.77
108	Muscat	686.62	66.15	69.08	71.69	66.08	50.08	40.15	35.85	36.77	44.15	56.92	73.77	75.92

Aggregated visitation values for calls not within hot-spot areas (remaining coumtry)

		PUD_YR_AVG	January	February	March	April	May	June	July	August	September	October	November	December
ID	Name
0	Country	285.85	41.62	29.85	26.15	20.77	12.15	12.85	14.54	12.62	15.23	28.0	38.69	33.38

Merge the hot-spot dataset and the remaining country record to a single table for further analysis.

Merged dataset, Visitation rateds for hot spots and remaining country

		PUD_YR_AVG	January	February	March	April	May	June	July	August	September	October	November	December
ID	Name
100	Salalah	201.54	15.15	11.85	14.46	13.77	9.85	9.69	20.00	27.69	28.31	17.46	19.54	13.77
101	Al Wasil	51.54	6.77	6.85	4.08	4.08	1.92	0.38	0.62	1.00	1.08	7.23	8.77	8.77
102	Sur	41.69	5.38	4.77	5.69	3.15	1.92	1.77	1.15	2.08	1.85	4.31	4.69	4.92
103	Al Buraimi	54.46	6.85	5.62	5.92	3.62	4.08	3.15	4.23	3.31	2.92	3.85	5.54	5.38
104	Sohar	23.77	1.31	2.54	2.15	2.62	2.85	2.08	1.69	1.85	1.08	1.38	2.46	1.77
105	Musandam	107.92	9.77	13.23	13.38	8.54	7.08	4.85	4.00	5.62	5.77	9.38	12.38	13.92
106	Nizwa	209.85	24.92	31.15	19.46	19.23	9.46	6.92	8.38	4.15	9.23	17.92	26.85	32.15
107	Dibah	77.15	10.77	9.92	6.15	6.92	4.85	3.92	2.31	2.62	2.92	8.62	7.38	10.77
108	Muscat	686.62	66.15	69.08	71.69	66.08	50.08	40.15	35.85	36.77	44.15	56.92	73.77	75.92
0	Country	285.85	41.62	29.85	26.15	20.77	12.15	12.85	14.54	12.62	15.23	28.00	38.69	33.38

Hot-spot areas with visitation percentage

	ID	Name	PUD_YR_AVG	Percentage
0	100	Salalah	201.54	11.60
1	101	Al Wasil	51.54	2.97
2	102	Sur	41.69	2.40
3	103	Al Buraimi	54.46	3.13
4	104	Sohar	23.77	1.37
5	105	Musandam	107.92	6.21
6	106	Nizwa	209.85	12.07
7	107	Dibah	77.15	4.44
8	108	Muscat	686.62	39.51

Calculations on visitation rates¶

The total visitation parameter for hot-spot areas is 1454
The total visitation parameter for the country of Oman is 1738
Hot-spot areas account for 83% of total visits in Oman

Oman Recreation Hot-Spots

MapView(layout=Layout(height='400px', width='100%'))

Data Analysis¶

In addition to hot-spot delineation, GIS tools allow planners to analyze the characteristics of hot-spot areas. In the section below, the visitation patterns in hot-spot areas will be analyzed to understand the overall visitor activity within these areas. This process can help planners to identify trends and patterns in visitor behavior and inform the development of strategies for managing and promoting these areas.

Cluster Analysis¶

Cluster analysis, also known as clustering, is a statistical method used to group data points into distinct clusters based on their similarities. In the case of hot-spot areas, cluster analysis can be used to group these areas based on the visitation patterns. By analyzing the number of visits, along with the month and year of the visit, it is possible to identify trends and patterns in visitor behavior and group hot-spot areas with similar patterns into distinct clusters. This can help planners to understand how visitors use these areas and identify which months have similar characteristics.

Cluster analysis can be useful in identifying patterns and trends in data and can provide insights into visitor behavior. It can also help to inform the development of strategies for managing and promoting hot-spot areas by identifying areas with similar characteristics and allowing planners to develop targeted approaches based on these similarities.

<seaborn.matrix.ClusterGrid at 0x26f70f5d0c8>

The results of the cluster analysis can be summarized as follows:

Two distinct periods define Tourism and Recreation in Oman:
1. The hot summer from May until September where the outdoor recreation is limited by the high temperature.
2. The mild period from October until April where the weather conditions allow outdoor recreation.
The hot-spot area clusters:
1. Nizwa
2. Shalala
3. Muscat
4. The rest of the hot-spot areas

The results of the cluster analysis can be summarized as follows:

There are two distinct periods that define Tourism and Recreation in Oman:

The hot summer from May until September, when outdoor recreation is limited by the high temperature.
The mild period from October until April, when the weather conditions allow for outdoor recreation.

The hot-spot area clusters are:

Nizwa,
Shalala,
Muscat, and the rest of the hot-spot areas.

Cluster analysis, or clustering, is a method used to group data points based on their characteristics. In this case, the hot-spot areas were grouped based on their visitation patterns, specifically the number of visits and the month and year of the visits. This allowed for the identification of distinct periods of tourism and recreation in Oman, as well as the grouping of hot-spot areas based on their visitation patterns.

Monthly Visitation Rate¶

	January	February	March	April	May	June	July	August	September	October	November	December
Name
Salalah	15.153846	11.846154	14.461538	13.769231	9.846154	9.692308	20.000000	27.692308	28.307692	17.461538	19.538462	13.769231
Al Wasil	6.769231	6.846154	4.076923	4.076923	1.923077	0.384615	0.615385	1.000000	1.076923	7.230769	8.769231	8.769231
Sur	5.384615	4.769231	5.692308	3.153846	1.923077	1.769231	1.153846	2.076923	1.846154	4.307692	4.692308	4.923077
Al Buraimi	6.846154	5.615385	5.923077	3.615385	4.076923	3.153846	4.230769	3.307692	2.923077	3.846154	5.538462	5.384615
Sohar	1.307692	2.538462	2.153846	2.615385	2.846154	2.076923	1.692308	1.846154	1.076923	1.384615	2.461538	1.769231
Musandam	9.769231	13.230769	13.384615	8.538462	7.076923	4.846154	4.000000	5.615385	5.769231	9.384615	12.384615	13.923077
Nizwa	24.923077	31.153846	19.461538	19.230769	9.461538	6.923077	8.384615	4.153846	9.230769	17.923077	26.846154	32.153846
Dibah	10.769231	9.923077	6.153846	6.923077	4.846154	3.923077	2.307692	2.615385	2.923077	8.615385	7.384615	10.769231
Muscat	66.153846	69.076923	71.692308	66.076923	50.076923	40.153846	35.846154	36.769231	44.153846	56.923077	73.769231	75.923077

Visitation rates per month.

The plot above shows:

That there is a trend in visitation rates in Oman where visitation drops significantly after April in all hot spots except in Shalala. Shalala is in it's peak while the rest of the hot-spots are in their lowest.
The three hot-spot areas that clearly differ in the cluster analysis have the significant role in the tourism industry in Oman. The three hot-spot areas are the only hot-spot areas of National importance. All other hot-spots have a regional significance.
Muscat is the main touristic destination in Oman. The low season extends from April until October. During the hot summer period, Muscat as a tourism hot-spot is not an attractive destination. The main development goal is to extent the tourism season in Muscat. So from the development perspective:
1. Investments in more hotels may not extent the high season.
2. Investment in the Muscat area should target on extending the tourism high period.

The plot above shows the trends in visitation rates in Oman, where:

Visitation drops significantly after April in all hot spots except in Shalala.
Shalala is in its peak while the rest of the hot spots are in their lowest.
The three hot spot areas that clearly differ in the cluster analysis have a significant role in the tourism industry in Oman.
The three hot spot areas are the only hot spot areas of national importance, while all other hot spots have regional significance.
Muscat is the main tourist destination in Oman, but during the hot summer period, it is not an attractive destination. Therefore, the main development goal should be to extend the tourism season in Muscat. This could involve investments in activities or attractions that are not weather-dependent, or marketing efforts to encourage visitation during the off-season.

Time series analysis¶

From the available data it is easy to examine the trend form 2005 to 2017.

Hot-spot visitation rates per year

There are several mild stones that have affected the flow to visitors in the past 15 years:

In 2008 the economic crises seems to have an impact to the industry but only for a limited period.
in 2011 there is a break point in the development of tourism in Oman.
Muscat lost the momentum in 2014.

Plots¶

In the section below the visitation distribution plots are available.

	poly_id	...	Total
0	0	...	5
2	3	...	5
4	4	...	10
5	5	...	8
6	6	...	14

	poly_id	...	Total
0	0	...	5
2	3	...	5
4	4	...	10
5	5	...	8
6	6	...	14

	poly_id	...	Total
0	0	...	5
2	3	...	5
4	4	...	10
5	5	...	8
6	6	...	14